Square footage adjustment

[alebrewer]

Segmenting your data based on specific market parameters alone isn't sufficient, you need to make sure you have enough sales to achieve reliable results. Additionally, the risk of being too granular is that you can always find (if you wanted to) the combination of parameters that yields the results that support your preconceived market trends.
I'm not saying your method is incorrect, I'm just pointing out the risks of using "micro" vs "macro" data.

You cannot - in my opinion - include all sales in a given geography as the subject's 'market'. They simply are not. The market for a $500k home is not the same market as that for a $200k home. Including those sales in the market analysis is, in my opinion, more egregious than the possibility of getting 'too granular'.

 

[Tom D]

So which is better to use, of which i have data access to and look at:

The metro area which includes me and 4 other counties. My city county. My zip code. My neighborhood. My sales search.

Depending on rural, sub, or urban. In my case big urban. The micro usually matches closely to the macro neighborhood. But that might not match the macro zip code, county or metro surrounding areas.

So in your macro of apples, peas, potatoes, you think the macro tells you enough about button mushroom prices.

 

[Mark K]

It's like high school when they made you show the way you got the answer - even though it was intuitively obvious. Now the lenders want you to show your work...

2 + 2 = 4 Show your work.

I got an F on an algebra test one time even though I had 100% correct answers. Is it my fault that I can do first year algebra in my head?

You cannot - in my opinion - include all sales in a given geography as the subject's 'market'. They simply are not. The market for a $500k home is not the same market as that for a $200k home. Including those sales in the market analysis is, in my opinion, more egregious than the possibility of getting 'too granular'.

Sales prices, overall, in this area are stagnant or falling, or increasing, depending on the price range of the home, location, and age, for starters.

Your basic, avg. price home ($300K here) is stagnant, extended market times compared to the past few years. Upper end homes, 2-3X avg., prices are falling. Anything in suburban or rural with 2-20 acres, prices are going crazy.

Using all of the sales in an area to determine price appreciation/depreciation is unreliable or lazy, maybe both.

 

[David S]

What about the time adjustments calculator? https://the-adjustments-app.streamlit.app/
Is that one similar to your actual grid adjustments?

I just tried with time adjustment. Couldn't interrupt what does the result mean. Just email it to you.

 

[weird_limbs]

Cool app, but I think it's only performing a single variable regression for GLA PPF, which - if that's the case - would tend to overstate the adjustment factor. My sensitivity analysis generally returns a factor of ~ $60-$80/foot for the market transactions I'm modeling, but this app suggested $140-$150. It's even more interesting when running the market conditions app. I'm in a pretty stable market, but this is what the model returned for market adjustments:

You are correct. The single variable regression was overstating the GLA adjustments. I modified the calculator to take into account the time adjustments and the site adjustments before running the GLA calculations. I think it works better now. Would you agree?: https://the-adjustments-app.streamlit.app/.
Thank you for the suggestion.

 

[alebrewer]

You are correct. The single variable regression was overstating the GLA adjustments. I modified the calculator to take into account the time adjustments and the site adjustments before running the GLA calculations. I think it works better now. Would you agree?: https://the-adjustments-app.streamlit.app/.
Thank you for the suggestion.

It's not showing the results of the analysis...

 

[weird_limbs]

It's not showing the results of the analysis...

There is no error message?? Weird.

 

[alebrewer]

There is no error message?? Weird.

nope...